Prostate adenocarcinoma (PCa) is a leading cause of death from cancer and there is no effective treatment for castration-resistant disease. This project is based on our recent findings that an actin-nucleating protein, Diaphanous related formin-3 (DRF3), is an inhibitor of the amoeboid phenotype in human PCa cells, and that chromosomal loss at the DRF3 coding locus occurs at high frequency in metastatic but not primary disease. Functional studies have demonstrated that DRF3 is a member of the EGF receptor (EGFR) network and is positioned at a signaling node that controls the transition from mesenchymal to amoeboid cancer cell phenotypes. The mesenchymal-amoeboid transition (MAT) has recently been identified as a dynamic phenotypic shift that can affect migration of tumor cells through matrices and reduce their sensitivity to protease inhibitors. DRF3 also inhibits the secretion of bioactive microvesicles capable of altering the tumor microenvironment and associates with cholesterol-rich lipid raft membrane microdomains. These data suggest that the DRF3 network may be sensitive to cholesterol- or fatty acid-targeting therapies and/or chemoprevention, which have shown promise in human studies and animal models. We will test the hypothesis that amoeboid properties arising from DRF3 loss results in a more aggressive phenotype. We further hypothesize that amoeboid behavior coincides with upregulation of lipid anabolism and that amoeboid tumor cells may become addicted to lipid-dependent pathways. Thirdly, we hypothesize that a distinct amoeboid signaling network exists and that components of this network may represent novel tumor biomarkers. The specific aims are: Aim 1. Determine the biological consequences of DRF3 loss in prostate cancer. The role of androgen, the androgen receptor, and ERBB receptor tyrosine kinase signaling in the amoeboid phenotype will be determined in the context of DRF3 silencing. We will assess whether DRF3 actively mediates the transition to the mesenchymal phenotype. The effect of DRF3 silencing on tumor growth, metastasis, and sensitivity to ERBB- and cholesterol-targeting therapy, and androgen ablation will be assessed. Patterns of DRF3 expression in human prostate cancers will be evaluated and correlated with clinical parameters. Aim 2. Identify the signaling network that mediates the mesenchymal-amoeboid transition in prostate cancer. A series of directed as well as unbiased experiments, evaluating perturbations in RNA, protein, and palmitoyl-protein networks under conditions of stable DRF3 silencing and EGFR activation will be employed to uncover the signaling network controlling the MAT in prostate cancer cells. Models will be constructed from these data and validated to identify a small series of informative indicators of the MAT. These markers will be used to determine whether the amoeboid phenotype can be detected in vivo using a state-of-the-art platform relevant to prognostic evaluation. These studies will provide new insight into the relationship between membrane dynamics, lipid metabolism and PCa progression to lethal disease.